Granite · 9B to 32B
Granite 4.1 30B
28.87 billion parameters across 64 layers. At four-bit precision the weights alone come to 16.23 GB, before any conversation is loaded. It uses grouped-query attention with 8 KV heads for 32 query heads, which cuts the cache to 25% of what full attention would need.
About Granite 4.1 30B
Granite 4.1 30B is a dense model with 28.87B parameters, built from 64 layers with a hidden size of 4096. Its repository was published on 2026-04-06. The license listed on Hugging Face is apache-2.0. It uses grouped-query attention with 8 KV heads for 32 query heads, which cuts the cache to 25% of what full attention would need.
Memory needed, by quantization
At an 8k context window. Quality is our rough ranking of how much the compression costs you: anything at or above Q5 is hard to tell apart from the original in normal use.
| Quantization | File size | KV cache | Total VRAM | Quality | What it costs you |
|---|---|---|---|---|---|
| Q8_0 | 28.57 GB | 2 GB | 30.97 GB | 99% | Lossless in practice. Use it when the memory is there. |
| Q6_K | 22.05 GB | 2 GB | 24.45 GB | 98% | Very close to Q8 for two thirds of the size. |
| Q5_K_M | 19.12 GB | 2 GB | 21.53 GB | 96% | The quality-first choice when Q6 will not fit. |
| Q4_K_M | 16.23 GB | 2 GB | 18.64 GB | 93% | The default. Best size-to-quality ratio for local use. |
| IQ4_XS | 14.28 GB | 2 GB | 16.69 GB | 91% | Importance-matrix 4-bit. Q4_K_S quality, smaller file. |
| Q3_K_M | 13.14 GB | 2 GB | 15.55 GB | 86% | Degradation starts to show. A way to fit one size up. |
| IQ3_XXS | 10.28 GB | 2 GB | 12.69 GB | 79% | Aggressive. Only worth it on very large models. |
What a longer conversation costs
Same model at Q4_K_M, only the context window changes. The cache grows in a straight line with every token in the window.
| Context | KV cache | Total VRAM | Fits in 8 GB | Fits in 12 GB | Fits in 24 GB |
|---|---|---|---|---|---|
| 4k | 1 GB | 17.64 GB | no | no | yes |
| 8k | 2 GB | 18.64 GB | no | no | yes |
| 16k | 4 GB | 20.64 GB | no | no | yes |
| 32k | 8 GB | 24.64 GB | no | no | no |
| 64k | 16 GB | 32.64 GB | no | no | no |
| 128k | 32 GB | 48.64 GB | no | no | no |
The "fits" columns allow for the roughly 0.8 GB Windows keeps for the desktop.
Which hardware runs Granite 4.1 30B
39 of 118 consumer devices run it at a quantization worth using, at an 8k context window. Another 32 can load it only by compressing the weights far enough to damage the model, marked with a warning below.
| Device | Memory | Verdict | Quantization | Used | Speed |
|---|---|---|---|---|---|
| GeForce RTX 5090 NVIDIA | 32 GB | Just fits | Q8_0 | 30.97 GB | 43.1 t/s |
| RTX PRO 6000 Blackwell NVIDIA | 96 GB | Runs great | Q8_0 | 30.97 GB | 43.1 t/s |
| RTX 6000 Ada Generation NVIDIA | 48 GB | Runs well | Q8_0 | 30.97 GB | 23.1 t/s |
| RTX A6000 NVIDIA | 48 GB | Runs well | Q8_0 | 30.97 GB | 18.5 t/s |
| Apple M3 Ultra 96GB Apple | 96 GB | Runs well | Q8_0 | 30.94 GB | 18.2 t/s |
| Apple M3 Ultra 256GB Apple | 256 GB | Runs well | Q8_0 | 30.94 GB | 18.2 t/s |
| Apple M3 Ultra 512GB Apple | 512 GB | Runs well | Q8_0 | 30.94 GB | 18.2 t/s |
| Apple M1 Ultra 64GB Apple | 64 GB | Runs well | Q8_0 | 30.94 GB | 17.8 t/s |
| Apple M1 Ultra 128GB Apple | 128 GB | Runs well | Q8_0 | 30.94 GB | 17.8 t/s |
| Apple M2 Ultra 64GB Apple | 64 GB | Runs well | Q8_0 | 30.94 GB | 17.8 t/s |
| Apple M2 Ultra 128GB Apple | 128 GB | Runs well | Q8_0 | 30.94 GB | 17.8 t/s |
| Apple M2 Ultra 192GB Apple | 192 GB | Runs well | Q8_0 | 30.94 GB | 17.8 t/s |
| Radeon PRO W7900 AMD | 48 GB | Runs well | Q8_0 | 30.97 GB | 17.6 t/s |
| GeForce RTX 4090 NVIDIA | 24 GB | Runs great | Q5_K_M | 21.53 GB | 35.1 t/s |
| GeForce RTX 3090 Ti NVIDIA | 24 GB | Runs great | Q5_K_M | 21.53 GB | 35.1 t/s |
| GeForce RTX 3090 NVIDIA | 24 GB | Runs great | Q5_K_M | 21.53 GB | 32.6 t/s |
| GeForce RTX 5090 Laptop NVIDIA | 24 GB | Runs great | Q5_K_M | 21.53 GB | 31.2 t/s |
| Radeon RX 7900 XTX AMD | 24 GB | Runs great | Q5_K_M | 21.53 GB | 28.4 t/s |
| RTX A5000 NVIDIA | 24 GB | Runs great | Q5_K_M | 21.53 GB | 26.7 t/s |
| Radeon RX 7900 XT AMD | 20 GB | Just fits | Q4_K_M | 18.64 GB | 27.4 t/s |
| Apple M4 Max 48GB Apple | 48 GB | Runs well | Q8_0 | 30.94 GB | 12.1 t/s |
| Apple M4 Max 64GB Apple | 64 GB | Runs well | Q8_0 | 30.94 GB | 12.1 t/s |
| Apple M4 Max 128GB Apple | 128 GB | Runs well | Q8_0 | 30.94 GB | 12.1 t/s |
| Apple M1 Max 64GB Apple | 64 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M2 Max 64GB Apple | 64 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
Show 35 more devices
| Apple M2 Max 96GB Apple | 96 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M3 Max 36GB Apple | 36 GB | Just fits | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M3 Max 48GB Apple | 48 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M3 Max 64GB Apple | 64 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M3 Max 96GB Apple | 96 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| Apple M3 Max 128GB Apple | 128 GB | Runs well | Q6_K | 24.42 GB | 11.3 t/s |
| NVIDIA DGX Spark 128GB NVIDIA | 128 GB | Runs well | Q4_K_M | 18.64 GB | 11 t/s |
| Apple M4 Pro 48GB Apple | 48 GB | Runs well | Q4_K_M | 18.61 GB | 10.2 t/s |
| Apple M4 Pro 64GB Apple | 64 GB | Runs well | Q4_K_M | 18.61 GB | 10.2 t/s |
| GeForce RTX 5080 NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 57.5 t/s |
| GeForce RTX 5070 Ti NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 53.7 t/s |
| GeForce RTX 5080 Laptop NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 46 t/s |
| GeForce RTX 4080 SUPER NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 44.1 t/s |
| GeForce RTX 4080 NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 42.9 t/s |
| GeForce RTX 4070 Ti SUPER NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 40.2 t/s |
| GeForce RTX 4090 Laptop NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 34.5 t/s |
| Radeon RX 9070 XT AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 32.8 t/s |
| Radeon RX 9070 AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 32.8 t/s |
| Radeon RX 7800 XT AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 31.7 t/s |
| Radeon RX 7900 GRE AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 29.3 t/s |
| GeForce RTX 5060 Ti 16GB NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 26.8 t/s |
| RTX A4000 NVIDIA | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 26.8 t/s |
| Radeon RX 6900 XT AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 26 t/s |
| Radeon RX 6800 AMD | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 26 t/s |
| Arc A770 16GB Intel | 16 GB | Runs great | IQ3_XXS ! | 12.69 GB | 25.9 t/s |
| GeForce RTX 4060 Ti 16GB NVIDIA | 16 GB | Runs well | IQ3_XXS ! | 12.69 GB | 17.2 t/s |
| Jetson AGX Orin 32GB NVIDIA | 32 GB | Fits, but slow | Q3_K_M | 15.55 GB | 10 t/s |
| Jetson AGX Orin 64GB NVIDIA | 64 GB | Fits, but slow | Q3_K_M | 15.55 GB | 10 t/s |
| Ryzen AI Max+ 395 64GB AMD | 64 GB | Fits, but slow | Q8_0 | 30.97 GB | 4.8 t/s |
| Ryzen AI Max+ 395 96GB AMD | 96 GB | Fits, but slow | Q8_0 | 30.97 GB | 4.8 t/s |
| Ryzen AI Max+ 395 128GB AMD | 128 GB | Fits, but slow | Q8_0 | 30.97 GB | 4.8 t/s |
| Radeon RX 7600 XT AMD | 16 GB | Runs well | IQ3_XXS ! | 12.69 GB | 14.6 t/s |
| GeForce RTX 3080 Ti NVIDIA | 12 GB | Runs great | IQ2_XXS ! | 9.33 GB | 75.2 t/s |
| GeForce RTX 3080 12GB NVIDIA | 12 GB | Runs great | IQ2_XXS ! | 9.33 GB | 75.2 t/s |
| GeForce RTX 5070 NVIDIA | 12 GB | Runs great | IQ2_XXS ! | 9.33 GB | 55.4 t/s |
How it compares
Other models in the same family or of a similar size, on a 12 GB card and a 24 GB card, each at the best quantization that fits.
| Model | Weights at 4 bits | GeForce RTX 3060 12GB | GeForce RTX 4090 |
|---|---|---|---|
| Granite 4.1 30B | 16.2 GB | does not fit well | 35.1 t/s at Q5_K_M |
| Granite 4.1 8B | 4.9 GB | 26.6 t/s at Q8_0 | 74.6 t/s at Q8_0 |
| Granite 4.1 3B | 1.9 GB | 66.4 t/s at Q8_0 | 185.9 t/s at Q8_0 |
| Gemma 4 31B Instruct | 17.6 GB | does not fit well | 35.5 t/s at Q4_K_M |
| Gemma 4 26B A4B Instruct | 14.5 GB | 37.6 t/s at Q4_K_M | 117 t/s at Q6_K |
| GPT OSS 20B | 11.3 GB | 65.7 t/s at MXFP4 | 180.1 t/s at MXFP4 |
What to buy to run Granite 4.1 30B
The cheapest hardware that runs it at a quantization worth using and a speed you would not resent, at an 8k context window.
Cheapest that works
Radeon RX 7900 XT
$650
Q4_K_M · uses 18.64 GB of its 20 GB · about 27.4 tokens/s
Best value
GeForce RTX 3090
$700 used
Q5_K_M · about 32.6 tokens/s · 936 GB/s
| Device | Price | Memory | Runs it at | Speed | |
|---|---|---|---|---|---|
| Radeon RX 7900 XT AMD | $650 | 20 GB | Q4_K_M | 27.4 t/s | Check price |
| GeForce RTX 3090 NVIDIA | $700 used | 24 GB | Q5_K_M | 32.6 t/s | Check price |
| GeForce RTX 3090 Ti NVIDIA | $800 used | 24 GB | Q5_K_M | 35.1 t/s | Check price |
| Apple M1 Max 32GB Apple · whole machine | $1,400 | 32 GB | IQ4_XS | 16.7 t/s | Check price |
| Apple M1 Max 64GB Apple · whole machine | $1,900 | 64 GB | IQ4_XS | 16.7 t/s | Check price |
| GeForce RTX 5090 NVIDIA | $2,200 | 32 GB | Q8_0 | 43.1 t/s | Check price |
| Apple M1 Ultra 128GB Apple · whole machine | $3,400 | 128 GB | Q8_0 | 17.8 t/s | Check price |
| Apple M3 Ultra 256GB Apple · whole machine | $5,599 | 256 GB | Q8_0 | 18.2 t/s | Check price |
Indicative prices reviewed 2026-09-01; used prices are marketplace typical. Options that cost more than a cheaper one with no more memory and no more speed are hidden. Price links are Amazon affiliate links. Change the standard, the context window or the budget in the buying tool.
Where it sits in the catalog
Source: ibm-granite/granite-4.1-30b on Hugging Face. Downloaded 340k times in the last month. Published 2026-04-06. Architecture figures are read from the repository's own configuration file, so they move when the model does.
Questions people ask about Granite 4.1 30B
How much VRAM does Granite 4.1 30B need?
At Q4_K_M with an 8k context window, Granite 4.1 30B needs about 18.64 GB: 16.23 GB of weights, 2 GB of KV cache and 0.41 GB of overhead. At the highest-quality file it needs about 30.97 GB.
What is the cheapest GPU that runs Granite 4.1 30B?
The smallest consumer device in our catalog that fits it is the Radeon RX 7900 XT with 20 GB, running it at Q4_K_M at roughly 27.4 tokens per second.
How much does a longer context window add for Granite 4.1 30B?
On the memory side, a 32k window adds 8.0 GB of KV cache on top of the weights, and the cache grows in a straight line with every token. The longest window the model supports natively is 128k tokens.